Source: raw/Patients_Look_You_Up_Before_They_Call_The_Videos_That_Get_You_More_Patients_AI_Avatar.md — “Patients Look You Up Before They Call: The Videos That Get You More Patients (AI Avatar)”, YouTube (youtube.com/watch?v=syt7Ea1aWYU), transcript fetched 2026-09-29; raw/You_re_Explaining_the_Same_Thing_3_Times_a_Week_Here_s_the_Fix_AI_Avatar_for_Doctors.md — “You’re Explaining the Same Thing 3 Times a Week? Here’s the Fix (AI Avatar for Doctors)”, YouTube (youtube.com/watch?v=JrS4PRDRIXc), uploaded 2026-09-16, 7:57, transcript fetched 2026-10-30 (HeyGen official channel, @heygen_official). Creator: presenter not named. The video ends “HeyGen is free to try. The link is below”, so treat it as vendor-adjacent. Platform: YouTube. Skill level: operator, no camera crew needed.

A placement playbook for medical practices: seven short AI-avatar videos, ordered by the point in the patient journey where each one is seen, from the first website visit to the post-visit review request. The presenter builds every example for a fictional clinic, “Quillhaven Family Care”, in HeyGen’s Video Agent, and draws a clear compliance line: the videos carry logistics and general information, never medical advice. The source has no performance data; its value is the video list and where each one goes.

Key Takeaways

  • Premise: “Before a new patient ever books an appointment, they’ve already looked you up and decided whether you feel like the right place.” Videos you make replace “whatever happens to be out there”.
  • Build method (HeyGen Video Agent): tell it what the video is for, paste the words you want said, answer its questions about the look and the presenter, and it builds the video “while you go and work on something else”.
  • Compliance comes first. Every video must be “general information only, not medical advice” and go through whoever reviews your practice’s materials before publishing. “HeyGen can help you make the video, but the compliance is on you.”
  • Keep clinical content with the provider. The prep example says to follow the provider’s instructions and call if unsure; the presenter: “I just want you to make sure the medical part stays with you. This video just handles the repeatable logistics.”
  • Cadence: you don’t need all seven this week. Start with the welcome video, “because that one works for you every single time somebody looks you up”, then add one a week.
  • The review ask is built into the journey. The last video goes in the post-visit email with “one small ask”, on the reasoning that “most people are happy to leave a review. You just got to ask them.”

The seven videos, in journey order

#VideoContentWhere it goes
1Welcome (~30 s)Who you are, who you see, how to bookTop of the website and “anywhere else someone online might find you”
2Front-desk FAQOne question, one answer (“Do you take my insurance?”), then a series: telehealth, parking, bringing kidsNot specified in the source
3What to expect at the first visitForms (portal or paper), what happens on arrivalAppointment-confirmation emails, “the exact moment” people wonder what to bring
4Prep (lab visit, procedure day)What to bring, who to call; logistics onlyReminder text the day before
5Service explainer (e.g. telehealth)Three simple steps; “built in one sitting from a four-line brief”Services page and patient portal
6Seasonal reminder (e.g. flu shots)How to book, where walk-in times are postedNot specified in the source
7Thank-you + review askVisit summary is in the portal; please leave a reviewPost-visit follow-up email

2026-10-30 addition: AI Avatar for Doctors (2026-09-16)

The setup episode of the same HeyGen healthcare series (“the video on building your own avatar and voice” the patient-journey video links to ^[inferred: the transcript names no title]). The presenter, Mary Lou of the fictional Quillhaven Family Care, is a video creator, not a provider (“That’s your job”). It builds the provider’s avatar and voice in one session and makes a ~20-second welcome clip. Premise: the explanation you gave three times this week (scan prep, the first visit, why the fasting window matters) becomes a video that “start[s] with typing.” No performance data.

Compliance line (same as the rest of the series). Anything published with the avatar is “general education and information, not advice for an individual patient”, and every video goes through the practice’s own review: “Clinical, legal, compliance, privacy, brand, whoever signs off at your practice signs off on these, too.”

Prepare three inputs, in this order:

  • Photo (decides how the avatar looks in every video): good even light on the face, looking straight at the camera, only you in frame. An existing headshot works, or HeyGen takes one with your webcam. Outfit and setting can change later without a new photo.
  • Video, about 45 seconds of you talking to the camera (“more if you’re feeling spicy”). It teaches HeyGen how your eyes, mouth and hands move, so energy matters more than being camera-ready. Sit centered, eyes on the lens, hands in frame, speak as you would to “a patient you like”, and pause between sentences: “the pauses are how it learns what you look like while you’re listening.” “Whatever energy you record is the energy your avatar brings to every video after this.” The presenter’s own training footage was shot on her phone.
  • Voice, recorded inside HeyGen in a quiet room.

Build steps (HeyGen web app):

  1. Avatar (left nav) → New avatar. HeyGen walks you through appearance → motion → voice.
  2. Appearance: upload the photo and name the avatar. Motion: upload the clip, or scan the code and record from your phone.
  3. Voice: read the script HeyGen supplies “the way you talk, not the way you perform it” (think of a voicemail to a long-time patient). When the script runs out, keep going with the words you say all day: the practice name, “call the front desk”, your most common instructions. “The clone learns the words you give it.”
  4. HeyGen builds a few voice versions to compare side by side; pick the closest and adjust.
  5. Verify identity: a short live-on-camera clip reading a code shown on screen. This is the consent step; it is “the reason no one else can build an avatar of your face.”
  6. First video: Home → create a new video → script → delivery → avatar → preview. Before rendering, HeyGen plays the voice reading your script; use the fix pronunciation option for clinical terms or your own name (“corrected once, and then it stays corrected”). Landscape for the website; generate.

Avatar governance, before anyone else touches it. “Your avatar will say anything someone typed.” Write rules on whose avatar it is, who may generate with it, what it may say, and that every video goes through the provider’s review before publishing. HeyGen lets you reset verification; nothing generates again until consent is re-recorded. “The control stays with you.”

Output of the session: a verified avatar and voice, a welcome clip to show the team, and written usage guidelines. The next episode turns it into a patient education video; see One Explainer Video per Repeated Patient Question (2026-10-30 addition).

Try It

  1. Write the welcome script first: who you are, who you see, how to book, about 30 seconds. Generate it and put it at the top of the website.
  2. Map the other six to messages you already send — the confirmation email, the day-before reminder text, the post-visit follow-up — so each video rides an existing touchpoint rather than a new channel.
  3. Start the FAQ series from your front desk’s most common question and add one question per video.
  4. Keep prep and reminder videos to logistics. Point anything clinical back to the provider, add the “general information only, not medical advice” disclaimer, and route every script through your practice’s review process.
  5. Add one video a week after the welcome video.

Open Questions

  • No performance data. The source gives no numbers on bookings, show rates, review volume or watch time, so the sequence is a recommendation, not a tested result. The video marketing measurement loop describes how to measure this yourself.
  • AI-presenter disclosure. Whether patients must be told the presenter is an AI avatar, and any healthcare-specific rules, are not addressed. The presenter’s own intro says the “doctor” is not real and was made in HeyGen, but the sample clinic scripts include no AI disclosure.
  • Placements for videos 2 and 6 are not stated.
  • Recording length differs across HeyGen tutorials. The doctors episode asks for about 45 seconds of training video; HeyGen Avatar V cites a 15-second motion clip. Whether the 45 seconds is a minimum, a recommendation for this flow, or a different model’s requirement is not stated.
  • Avatar use policy template. The doctors episode says to write usage rules but gives no template, and does not say whether HeyGen supports per-user permissions on a workspace avatar.
  • Presenter and relationship to HeyGen are unconfirmed; the closing call to action suggests a HeyGen-produced video.

See also: later videos in the same HeyGen healthcare series cover the dental version, the PDF-handout route, and a full build of the telehealth explainer (video 5).